warpstorm
Yumbo? Yumbo!
Yes, TheNiceOne, if you say anything good about Civ3 or Firaxis, you are a fanboy.
Another Plea for the Innocent Numbers!
- Thread starter Agamemnus
- Start date
Agamemnus
Warlord
Hehe. The RNG functions correctly for battles, not bombard (if the bombardment value is too low, that doesn't mean you should get a hit/no hit like this: 1-0-1-0-1-0-0-0-0-0-0-0-0-0-0-1-0-1-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-1-0-1-0-0-0-0-0-0-0-0-1-0. Rather, it should be more like this: 1-0-0-0-0-1-0-0-0-1-0-0-0-1-0-0-0-1-0-1-1-0-0-0-1-0-1-0-0-0-0.) If they're calculated the same way, then I'm wrong.
Having studied probability, and random number generation, the first example you give looks more truly random. 
The nice one.
I respect your opinion, my post wasnt adress to you specificaly, but when you said a pickman should stand a chance against modern armor then i prefer to dont argue with you on that point. i hope you understand why.
I read something about ignorant in your post, i have a diplom in analytical chemistry and i did a lots of mathemathics to get it so i dont think the hat fit on me.
As you know many poeple here post fanzy calculation about this game and i beleive most of them will fail to an expert analysis. I find pretty strange that Mike B. never argue on this kind of hypothesis.
I respect your opinion, my post wasnt adress to you specificaly, but when you said a pickman should stand a chance against modern armor then i prefer to dont argue with you on that point. i hope you understand why.
I read something about ignorant in your post, i have a diplom in analytical chemistry and i did a lots of mathemathics to get it so i dont think the hat fit on me.
As you know many poeple here post fanzy calculation about this game and i beleive most of them will fail to an expert analysis. I find pretty strange that Mike B. never argue on this kind of hypothesis.
Agamemnus
Warlord
Ok, fine. How about then we just kill the bombard random factor. Make it 0-1-0-1-0-1 or 1-1-1-1-1-1-1-1-1-1. 
Originally posted by Agamemnus
Ok, fine. How about then we just kill the bombard random factor. Make it 0-1-0-1-0-1 or 1-1-1-1-1-1-1-1-1-1.
The best way to make your bombardement effective is to lower defensive bonus agaisnt bombardement for building and citizen. Firaxis increase this value from 4 to 16 throught patch. Lower it back to 4 into editor under general setting. Since when an undefended city should have greater defense ( 16) then the attacking power of bomber ( 8). Maybe same funny reason why a pickman should stand a chance ag modern armor.
If you want this game make sens, you need to modified it.
warpstorm
Yumbo? Yumbo!
The thing you are missing on the bombard is the fact that there might not be a target of the chosen type. Soren has posted on this repeatedly in the past (but maybe it was on that other forum). When you bombard, first it picks a target type (building, population, or unit with more than one hit (unless lethal bombing is set)) then if there is something of that type to target the attack takes place with normal probability, otherwise it automatically misses.
The statistical analyses of this game has been checked and rechecked numerous times. There are tools out there to give you the odds on any given attack. In general, the odds are lower than you think (unless you actually understand the rules and statistics).
The statistical analyses of this game has been checked and rechecked numerous times. There are tools out there to give you the odds on any given attack. In general, the odds are lower than you think (unless you actually understand the rules and statistics).
sabo
My Ancestors were Vikings
Originally posted by Kilroy
Use the misunderstood, seldom used, underrated search feature to find out
alexander dumas
Chieftain
Tassador,
From a post you made on the first page, should I take it that you are advocating having a bunch of numbered balls attached to the computer so that we can get truly random numbers? Also, you were right in pointing out to the Nice One that he shouldn't call strangers ignorant, but to then go and pull a diploma card was sad. Truly sad.
I'm also slightly frustrated that this is just another plea to mod the game for the basis of "realism." This thread should be over.
From a post you made on the first page, should I take it that you are advocating having a bunch of numbered balls attached to the computer so that we can get truly random numbers? Also, you were right in pointing out to the Nice One that he shouldn't call strangers ignorant, but to then go and pull a diploma card was sad. Truly sad.
I'm also slightly frustrated that this is just another plea to mod the game for the basis of "realism." This thread should be over.
How nice. Personally, I have a B.S. that's actually in mathematics (and two more in physics and astronomy) and am well on my way to earning my PhD in theoretical high energy physics - a field using vastly more and more sophisticated mathematics than chemistry. To give one example, one of my recent classes spent almost a full month discussing the ins and outs of PRNGs. Some of that information I will pass on below.
The fact of the matter is, most PRNGs are pretty decent. Even the one incorporated into most C compilers as the rand() function - most compliers apparently use what was written into the standards as an example of the absolute minimum that would fulfill the standard - produces a distribution that (up to the effects of a computer's finite precision, which are negligible for this purpose) is good, but not perfect. Though the distribution of single PRNs is indistinguishable from a true uniform random distribution, a series of PRNs exhibits measurable (with a very large sample size) correlations. This would POSSIBLY be detectable in CivIII by amassing many thousands upon thousands of battle results and looking, not for things like "Do spearmen beat modern armor too often?", but for things like, "If a spearman beats a modern armor in the first battle of a sequence of tests, is the probability that the attacker will win battle 21 in that series too high?" There is NO WAY it could be detected with anything less than a concerted, multi-month, sophisticated effort, since we don't have access to the PRNG itself, just the battle results.
Other, better, PRNGs exist that are "perfect" in any practical sense. In at least one case, there has been a $1000 prize announced for anyone who can find a nontrival statistical test the PRNG fails. This prize has stood unclaimed for more than a decade.
For a useful reference, see Numerical Recipes in C, in particular sections 7.0 and 7.1.
Hopefully this has helped educate some people about the realities of PRNGs.
Would you accept a Bachelor's degree in Chemistry & Biology, with a minor in Physics?Originally posted by Tassadar
Studying is not enough to make you an expert, you need the diplom
Edit: I see Beamup has me beat out in the applicability of degree.
So I defer to his statements on the RNG (or Pseudo RNG as he points out ), which, BTW, I agree with.exiledviking
Chieftain
Hi, this is my first post here, but I've been reading the boards for a few months.
Who says a pikeman (or indeed any other unit, warrior included) can't occasionally beat a tank unit??? During WW2, Finnish troops made the infamous Molotov cocktails and crashed them on the exhaust area of passing tanks. Surely, that would be possible for almost any unit?
Also, getting near a tank and then jumping up onto it would probably allow some form of (rare) victory.
Who says a pikeman (or indeed any other unit, warrior included) can't occasionally beat a tank unit??? During WW2, Finnish troops made the infamous Molotov cocktails and crashed them on the exhaust area of passing tanks. Surely, that would be possible for almost any unit?
Also, getting near a tank and then jumping up onto it would probably allow some form of (rare) victory.
Welcome, exiledviking!
Indeed, this type of explanation is often given by those who feel the results of the RNG are acceptable. But those who feel that a tank should always beat a spearman are not swayed by such talk, nor by discussion of such things as "game mechanics" and "balance".
Agamemnus
Warlord
Nonono. I don't see where the imbalance is of crushing other civs who have, for instance, cavalry and riflemen, if you have Tanks.
I don't like all these artificial technology barriers in the default rules. There isn't a randomness slider for battles in CivEdit. There should be a slider, either in-game, or CivEdit.
Even if a spearman jumps on a tank passing thru a road (ambush), you can't ambush a whole column of Tanks, and the units in Civ3 don't just represent one of something. For instance, a laborer could represent 10,000 citizens.
I don't like all these artificial technology barriers in the default rules. There isn't a randomness slider for battles in CivEdit. There should be a slider, either in-game, or CivEdit.
Even if a spearman jumps on a tank passing thru a road (ambush), you can't ambush a whole column of Tanks, and the units in Civ3 don't just represent one of something. For instance, a laborer could represent 10,000 citizens.
Gastric ReFlux
Dispatch our Tech Staff
- Joined
- Jun 25, 2002
- Messages
- 717
Or, alternatively, that spearman doesn't just represent any old spearman, but one that's been around for maybe 3,000 years, has probably received extensive spearman guild training, swapped a few stories with other spearmen buddies, and could probably moonlight as an assassin who kills with a feather duster.
It's as Padma said above. Some can't accept it, others can.
Zachriel
Kaiser
There is a slider. Increasing unit hitpoints in the editor will reduce the effect of "randomness."
Agamemnus
Warlord
Or, alternatively, that spearman doesn't just represent any old spearman, but one that's been around for maybe 3,000 years, has probably received extensive spearman guild training, swapped a few stories with other spearmen buddies, and could probably moonlight as an assassin who kills with a feather duster.
LOL
So that's why we need a slider, man!
Zachriel: But what if I wanted to play with the same stats? And what about CivIII multiplayer or GOTM or something similar that uses standard rules, there likely won't be an option to use a mod in it, but a slider is like 'Cultural Victory off', 'Emperor difficulty', 'civ=Greece', etc.
LOLSo that's why we need a slider, man!
Zachriel: But what if I wanted to play with the same stats? And what about CivIII multiplayer or GOTM or something similar that uses standard rules, there likely won't be an option to use a mod in it, but a slider is like 'Cultural Victory off', 'Emperor difficulty', 'civ=Greece', etc.
Beamup
I like to discuss with poeple like you, if i understand well ( because my english is not perfect, i am french canadian), you said about exactly what i said; rng look random but they are not. As you said with many trival we can find a pattern or even more predict the next number. It happen here in quebec, in a casino, they use a programm to generate number but a very intelligent guy was able to predict the number, he win 400 000 $. It was a major crisis here. and subsequently programer notice it was a default in the clock ( time ) that create this possibility.
So since that, i think they use fluctuation in power supply to generate synthetic random number, does anybody know how to manipulate power fluctutation ? Then he would be able to predict any number generated by this kind of programm.
But to keep it simple, do you realy think a pickman have a chance against M1a1 abrham thanks ? me i dont think so
I like to discuss with poeple like you, if i understand well ( because my english is not perfect, i am french canadian), you said about exactly what i said; rng look random but they are not. As you said with many trival we can find a pattern or even more predict the next number. It happen here in quebec, in a casino, they use a programm to generate number but a very intelligent guy was able to predict the number, he win 400 000 $. It was a major crisis here. and subsequently programer notice it was a default in the clock ( time ) that create this possibility.
So since that, i think they use fluctuation in power supply to generate synthetic random number, does anybody know how to manipulate power fluctutation ? Then he would be able to predict any number generated by this kind of programm.
But to keep it simple, do you realy think a pickman have a chance against M1a1 abrham thanks ? me i dont think so
You know this old fashion way of killing thank with molotov cocktail is more a legend then anything else, even more the soldier Ryan( movie) sticky bomb.
With those kind of primitive technology the worst you can do is disable thank track ( stop mobility). But even then the thank machine gun are able to fire. But this was with ww 2 thank, now modern armor, or m1a1 abrham ( 1980 technology) are able to sustand rocket attack, and even against stand a chance ag airplane bomb. They got explosive uranium shield, they got auto aim machine gun, they got homing missile, please dont make me beleive a medeival pickman can defaet this war machine.
Edit; Alexander Dumas
you know it is imposible to live without programmed a.i. simulation battle. We both agree with that, but in civ 2 i know that i will defeat a pickman everytime with a tank ( which was realistic) but now in civ 3 a caveman got a statistical chance to defeat a modern armor, which is totaly illogical, unrealistic,imposible and out of common sens.
Similar threads
